Comparative typology in six european low-intensity systems of grassland management

(1)

CORRECTED

PROO

F

C H A P T E R S E V E N

Comparative Typology in Six

European Low-Intensity Systems

of Grassland Management

Rafael Caballero,1Jan A˚ge Riseth,2,3Niklas Labba,3Ewa Tyran,4

Wieslaw Musial,5Edyta Molik,6Andrea Boltshauser,7

Pius Hofstetter,8Anne Gueydon,9Norbert Roeder,10

Helmut Hoffmann,9Manuel Belo Moreira,11

Inoceˆncio Seita Coelho,12Olga Brito11_{and A´ngel Gil}1

Contents

1. Introduction 353

2. Presentation of Study Areas 355

2.1. Northern Sapmi, Fennoscandia 355

2.2. Tatra mountains, Poland 358

2.3. UNESCO Biosphere Entlebuch, Switzerland 359

2.4. Bavaria, Germany 359

2.5. Baixo Alentejo, Portugal 360

2.6. Castile-La Mancha, Spain 361

3. Material and Methods 361

3.1. Main criteria and indicators 362

3.2. Management units 367

3.3. Sampling process 368

1

Centro de Ciencias Medioambientales, CSIC, Madrid, Castile-La Mancha, Spain

2

NORUT Ltd., Troms, Norway, Northern Sapmi, Scandinavia

3

Sa´mi Institute, Kautokeino, Norway, Northern Sapmi, Scandinavia

4

Department of Agribusiness, Agricultural University of Krakow, Tatra Mountains, Poland

5

Department of Agricultural Economics and Organization, Agricultural University of Krakow, Tatra Mountains, Poland

6

Department of Sheep and Goat Breeding, Agricultural University of Krakow, Tatra Mountains, Poland

7

UNESCO Biosphere Reserve Entlebuch, CH-Schupfheim, Entlebuch, Switzerland

8

Schupfheim Agricultural Education and Extension Center, CH-Schupfheim, Entlebuch, Switzerland

9

Lehrstuhl fu¨r Wirtschaftslehre des Landbaues, Technische Universita¨t Mu¨nchen, Bavaria, Germany

10

TUM Business Scholl, Environmental Economics & Agricultural Policy Group, Technische Universita¨t Mu¨nchen, Bavaria, Germany

11

Instituto Superior de Agronomia, Technical University of Lisbon, Baixo Alentejo, Portugal

12

Instituto Nacional de Investigac¸a¨oo Agra´ria e Pescas, Ministe´rio da Agricultura, Desenvolvimento Rural e Pescas, Lisbon, Baixo Alentejo, Portugal

(2)

CORRECTED

PROO

F

4. Results 370

4.1. Land uses 370

4.2. Size of farm-holding, land prices, and grazing fees 372

4.3. Institutional economics 375

4.4. Institutional and legal frameworks 379

4.5. Forage deficit 381

4.6. Grazing infrastructure 385

4.7. Labor 388

4.8. Productivity estimates 390

4.9. Economic performance 395

4.10. Grazing management and trends 401

4.11. Main limiting factors 404

4.12. Interface to biodiversity 406

5. Discussion 408

References 414

European biodiversity significantly depends on large-scale livestock systems with low input levels. In most countries forms of grazing are organized in permanent or seasonal cooperations (land-owner/land-user agents) and covers different landscape such as alpine areas, forest, grasslands, mires, and even arable land. Today, the existence of these structures is threatened due to changes in agricultural land use practices and erratic governmental policies. The present chapter investigates six low-input livestock systems of grassland management with varying degrees of arrangements in different European countries and landscapes. These large-scale grazing systems (LSGS) are rein-deer husbandry in Northern Sapmi (Fennoscandia), sheep grazing in the Polish Tatra mountains, cattle grazing in the Swiss and German Alps, cattle, sheep, and pig grazing in Baixo Alentejo, Southern Portugal, and sedentary sheep grazing in Central Spain. These systems showed very heterogeneous organizational patterns in their way of exploiting the pastoral resources. At the same time, these LSGS showed at least some of the following weaknesses such as poor economic performance, social fragility, and structural shortcomings for proper grazing management. Lack of proper mobility of herds/flocks or accession to specific grazing grounds can be a cause of environmental hazards. The sur-veyed LSGS are mostly dependent on public handouts for survival, but succes-sive policy schemes have only showed mixed effects and, in particular study areas, clear inconsistencies in their aim to stop the general declining trend of LSGS. This research assumed that detailed system research may open the way for better-focused policy intervention, but policymakers need to take advantage of this period of support to push ahead for reforms. Recent European Union (EU) guidelines (2007–2013) on Rural Development Policy (RDP) and its operative scale of high nature value (HNV) farmland can easily fit the structure and functions of low-input grazing systems and LSGS.

(3)

CORRECTED

PROO

F

CORRECTED

PROO

F

1. Introduction

Large-scale extensive livestock systems in Europe represent large chunks of European land relative to the size of the business operations. As large-scale systems, they may represent links with nature values at the landscape level, and as extensive systems, they may represent low input and low value of production relative to the size of the business at the farming level. These large-scale grazing systems (LSGS) are mainly located in the most remote and less favored areas (LFAs) with harsh environmental and sometimes difficult social conditions. A small part of the rural population stands to make a living by maintaining traditional grazing practices, which, in turn, shaped the environment.

As being located in mostly developed countries, these systems have faced two main threats: intensification and abandonment (the most extreme form of extensification). In the first case, the harsh environments have limited the impact. In the second case, some studies have pointed out the risk of abandonment in the LFAs of the European Union (EU) (Baudry et al., 1996; Caravelli, 2000; Garcia-Ruiz et al., 1996; Muller, 1996; Zervas, 1998), but a pan-European coordinate socioeconomic research on the viability of LSGS is still lacking. Farming systems thought to satisfy ecologi-cal sustainability objectives must be economiecologi-cally attractive to farmers, if they are to be voluntarily adopted and continued (Dobbs, 2004). However, only 6 of the 22 studies on extensification of European livestock systems detailed in a review paper (Marriot et al., 2004) collected data on animal performance and only two individual studies showed some indicators on economic performance. None of these studies was previously coordinated at the European level.

Maintenance of LSGS may thus be dependent on the fact that marginal-ization of agriculture, undermining viability of rural communities, does not go so far. In turn, LSGS may manage to fill in gaps created by a declining intensity of land use. In fact, the LFAs of the EU-12 represented some 56% of the EU’s total surface area, and contained much of the high nature value (HNV) farmland (Brouwer et al., 1997). Left to their own or under insensi-ble schemes of policy support, the abandonment threat can be more preju-dicial than the intensification threat (Atance et al., 2000; Kristensen et al., 2004; Vicente-Serrano et al., 2004). Both threats, however, may derive similar effects: disappearance of potential economic, environmental, and social values (Angelstan et al., 2003; Donald et al., 2002; Krohmer and Deil, 2003; Loumou and Giourga, 2003; Tucker and Heath, 1994; Waldhardt et al., 2004). It is common ground to highlight the importance of the agronomic and environmental services (pollinators, biological pest control, cultivated plants, and wild relatives, and so on) provided by these relatively undisturbed ‘‘natural ecosystems’’ (Hillel and Rosenzweig, 2005). It is less

(4)

CORRECTED

PROO

F

common to change the arrow of causality and posing the question on how these truly LSGS are going to survive and continue to provide their potential assets.

European extensive systems of grassland management, notwithstanding their ample variation in environmental and structural components, face an encounter with modern farming or farming intensification. Can they sur-vive with tactical concessions to modernity? Do they share some qualities that helped them to survive? Are they adept at anticipating or adapting to changes? Are their stakeholders shrewd managers of their assets? Are they good advertisers of their cultural, economic, or environmental utility? Are these systems able to integrate new values and functions to their products? These are some questions to which a typology of policy relevance may provide some answers. This research will argue that European policy intervention can be devised at the space scale of LSGS and HNV farmland. Structural and social constraints, as well as potential environmental assets, are linked to specific systems. Sensible policy schemes can only be devised after untangling these constraints.

In the following chapter, we sum up the results of a parametric analysis of six European study areas. This study was conducted within the EU-funded research project ‘‘Landscape Development, Biodiversity and Coop-erative Livestock System’’ (Caballero and Ferna´ndez-Santos, 2004; LACOPE, 2002). These six study areas, and their respective LSGS, cover a wide range of different ecological, social, and economic conditions and exhibit different adaptations of the grazing system. The investigated large-scale extensive grazing systems are representative for some of the most widespread types of this kind of agricultural land use. One study focused on the reindeer grazing system of the boreal-alpine biogeographical region (Northern Sapmi); three LSGS represented different mountainous sys-tems (Tatra Mountains in Poland, Bavarian Alps, and Swiss Alps). The remaining two covered outstanding Mediterranean systems: the open fields of Campo Branco and the surrounding Montado system in Portugal and the cereal–sheep system in Spain.

Extensification is the process of reducing fertilizer inputs, management intensity, and stocking rates at the farm level and is central to sustainable rural policies. However, typology research in the LFAs is fragmented and exten-sification studies should adopt an approach that will allow their results to be applied throughout Europe (Marriot et al., 2004; Strijker, 2005). Successfully decoupling payments from production while maintaining HNV farming systems represent a severe challenge to the Common Agricultural Policy (CAP) of the EU. Studies are needed across a range of HNV areas in all of Europe’s biogeographical zones (Beaufoy et al., 2003). The main objective of this research was to assess whether some typology categories or common features can be drawn from the data of the six study areas or divergences between systems are perceptive for most headings and indicators.

(5)

CORRECTED

PROO

F

2. Presentation of Study Areas

All study areas have in common that they represent pastoral systems under harsh environments. A significant part of the biodiversity depends on open land and the maintenance of particular pastures. To dismiss pastoralism as a backward pursuit, an embarrassment to notions of modernization, is to put aside a proven response to harsh environments. In most study areas, except Baixo Alentejo (Portugal) and to a lesser extent Entlebuch (Swiss Alps), pastoralists are not the owners of the land. This fact may represent problems of mobility of herds/flocks or access to pastureland. Full pastoral operations are carried out in the management unit (MU). This MU may encompass more than one farming unit of seasonal grazing use. Four of the investigated systems (Sa´mi reindeer management and the three alpine sys-tems in central Europe) are still showing a pronounced seasonal migration pattern. The distance can easily exceed 200 km in Northern Sapmi, while it is rather short in the three alpine systems. The two Iberian systems are sedentary, albeit not devoid of mobility and access problems, especially in the cereal–sheep system of Castile-La Mancha (Spain). Key figures of the six study areas are shown in Table 1. Geographical location of the seven LACOPE study areas is depicted in Map 1. The study area of Connemara (west of Ireland) was not integrated in this report.

2.1. Northern Sapmi, Fennoscandia

These LSGS take up a large tract of the northern part of the Scandinavian peninsula (Northern Fennoscandia), encompassing land of Norway, Swe-den, and Finland. Reindeer management culture by Sa´mi herders is well entrenched in the area. National differences exist in historical background (Sandberg, 2006), herder production strategies (Riseth, 2000, 2003, 2006), as well as in the national legal and subsidy systems. In Sweden and Norway, reindeer management, with a couple of regional exceptions, is culturally and ethnically connected to Sa´mi people, while in Finland it is open to everyone and mainly a side industry to agriculture. Full seasonal migration, short-distance migration, and stationary patterns of reindeer herding can be envisaged, the first most common in Norway and Sweden and the latter in Finland.

In Norway, the husbandry unit is the base for most subsidies. Husbandry unit leaders, by cultural tradition, are usually concession holders. The concession model, dated from 1978, is the legal foundation for awarding subsidies. Within husbandry units, other right holders, apart from the leader, can be herders with subsidy allocation rights. In the western Finnmark area

(6)

CORRECTED

PROO

F

Table 1 K e y figures of the regional agriculture N o rt h e rn Sa pm i a T at ra M ou nt ai ns b Ent leb uch c Ba v a ri a d Bai x o A le n tejo e Cas ti le -La Ma n cha f Total acreage of LSGS per study area (ha) g 42,000,000 2500 g 7000 g 61,000 g 220,000 6,000,000 Grassland (t/ha dry matter) 1–1.5 2.5 (mountain), ₄(valley) 1.5 (mountain), _7.5 (valley) 7 (foothills) 0.8 2 Wheat grain (t/ha) – – – – 1 2.2 Orientation of the livestock production on the LSGS Reindeer (meat-oriented) Sheep (milk and meat-oriented) Mainly heifers Mainly heifers Meat-oriented flocks (sheep, black pig, cattle) Sheep (milk and meat-oriented) a S yla nd et al . (20 02). Rein driftsfor valtningen, 2005. b Statistical Yearbo ok (20 02). Data onl y for Tatra National Park. c Re gional managemen t Bio spha ¨renreservat Entl ebuch (2002) . d LBA (2002) ; Agrarg ebiet: Alpenv orland; meadows cut four to five times. e INE (20 01, 2002) and de Sequeira (19 88) for the producti vity figures. fCaballero (2001) . g Tot al acre age of th e grazi ng syste ms unde r collective form.

(7)

CORRECTED

PROO

F

Biogeographical regions

Alpine Anatolian Artic Atlantic Black sea Boreal Continental Mediterranean Pannonian Steppic

Study areas

N

1. Northern Sapmi, Fennoscandia 2. Connemara, Ireland 3. Tatra, Poland 4. Bavaria, Germany 5. Entlebuch, Switzerland 6. Castile-La Mancha, Spain 7. Baixo Alentejo, Portugal

National borders 0 1000 2000 3000 4000 5000 Kilometers Map 1 Bi o geog ra p h ic a l re g ions of E u ro p e a n d loca tion (ho t sp o ts) of L A C O P E study a rea s. S our ce :http :/ /d at a se rvice .e ea .e u . in t/ da ta se rvice /meta detal ls .a ssp ?ta b le= Bi o g e o 01 & j=1 .

(8)

CORRECTED

PROO

F

districts, 241 husbandry units, 1279 reindeer owners, and 84,200 reindeers are in the area (year 2003–2004). The subsidies are bound to the concession holder who has to produce a certain minimum amount of meat to qualify. Reindeer herders with no concession are not qualified for subsidies and with little incentive to get into the reindeer industry. Reindeer families are represented for one husbandry leader.

The Swedish and Finish subsidy systems are not based on concession holders. In Sweden, the system is based on reindeer owners. Any owner with a minimum level of meat production counts in the official statistics. In Finland, the subsidy system requires husbandry masters to reach a mini-mum of 80 animals to qualify. Reindeer owners, with reindeer herding

as their main source of living, receive a headage payment of 22€ per animal.

In Sweden and Finland, one reindeer husbandry family may encom-pass several husbandry master or doallu (household). Regional data for Sweden (northern Norbotten la¨ns) average 332 husbandry masters, 1249 reindeer owners, and 56,522 reindeer in the area. Regional data for Finland (Ka¨sivarren paliskunta area) encompassed 128 husbandry masters, 168 reindeer owners, and 10,000 reindeer in the area (Paliskuntain, 2004).

2.2. Tatra mountains, Poland

The LSGS in the Polish Tatra mountain (Carpathian region) are strongly linked to milk sheep. These sheep stay from late autumn to mid-spring on the lowland farms and graze clearings in the mountain forests or areas above the timberline (alps) in the summer for roughly 160 days under care of a flock master (baca). Usually, one baca, which is a small sheep farmer, gathers the sheep of other small farmers (gazdas) and takes them together to the alps. One average baca flock (some 300–350 sheep) can be composed of the flocks of some 20 sheep of different owners. Sheep flocks under baca’s care are allocated to several clearings (around 12 clearings per baca of an average size of some 5 ha) in the alpine forests.

Some 75% of the clearings are privately owned and 25% is public (Tatra National Park) land. The clearings have several owners who claim property but do not have proper legal documents. Similarly, bacas cannot claim for subsidies in the alps as they do not have proper renting documents. As a whole result, devising and implementing proper policy schemes and incen-tives for moving sheep to the natural pastures in the alpine and subalpine zone (alps) cannot be properly established without an overhauling of the legal and institutional framework. Currently, farmers can only receive subsidies for land and sheep they own on the lowland farms.

Although sharing shepherding for the summer season is traditional, cooperation between different stakeholders (bacas, gazdas, and landowners) is difficult to manage, people are independent and difficult to be engaged for any form of cooperation. Some 20 years ago, sheep were moved to distant

(9)

CORRECTED

PROO

F

clearings (up to 100–150 km). These clearings are currently underused due to difficulties in transportation and the low profitability. There is a need of keeping traditional ways of milk processing, but also to fulfill EU sanitary requirements. From mid-October to early May, individual sheep flocks are moved to lowland farms and fed indoor with hay produced in summer, while sheep flocks in the clearings.

2.3. UNESCO Biosphere Entlebuch, Switzerland

In the UNESCO Biosphere Entlebuch (canton Lucerne) there are 211 units of alpine pastures (Hofstetter et al., 2006a) and 1015 (AfS, 2007) farm units in the valleys in the year 2005. Some 20% of the farm units of the valleys shared also one alpine unit, moving animals to the latter in the summer for some 110–130 grazing days, depending on the altitude, exposition, and grazing management (Hofstetter et al., 2006b). Boarded (external) animals stay an average of 110 days with small variation. Owned animals stay a mean of 130 days with larger variations, depending on whether and how much the owner provides alpine pastures’ hay to the herd.

Most summer pastures are located in the prealpine area (67% between 1200 and 1400 m) and a small part in the high alpine area (up to 2500 m). The dominant grazing lot is heifers from the lowland farms. Most alpine units (72%) have only one stable and 28% more than one when allocated pastures of different altitude.

Often, the area of the lowland farm unit in Entlebuch with an average of 14 ha (Hofstetter et al., 2006a) is small compared with the area of the alpine unit (mean of 57 ha). As a consequence, private owners or tenants of alpine units cannot properly stock the alpine units only with their own animals, and thus external animals are added to the mixed herd. Owners of these external animals pay a grazing fee to the owner/tenant of the alpine unit (LBL, 2004a). These grazing fees, together with the subsidies, are the main sources of income of most alpine units over the summer season. In the alpine units of Entlebuch, most external animals are heifers and sheep from the canton of Lucerne (Office for Agriculture and Forest, 2004).

2.4. Bavaria, Germany

The Bavarian study area covers the German part of the Alps and the foothills

in their vicinity. In the year 2005,50,000 cattle grazed on rough pastures

in this area (Miller, 2006). Heifers are the dominant livestock species on these rough pastures, which are mainly alps, or less frequently grazed moor-lands in the foothills. Over 40% of the area used by this grazing system is under some form of cooperative livestock management or cooperative livestock system (CLS). This makes the Bavarian study one of the remaining strongholds of CLS, which were fairly widespread in Germany until the

(10)

CORRECTED

PROO

F

nineteenth century. For convenience, all CLS in the Bavarian study areas will be labeled Allmende. Typically, several lowland dairy farmers participate in the use of one Allmende on which they raise their replacement heifers during the vegetation period. Despite the awarded incentives, the intensity of land use declined within the last 20 years. The number of active mem-bers, the ones who send animals to the pastures of the Allmende, not only dropped from 27 to 17—or 2% per year—for the average Allmende, but also the number of animals using these pastures declined. The mean size of the upland Allmende units in the sample was some 400 ha. The average unit is more or less evenly divided between forest and grazing land. A shared characteristic of the Bavarian and Swiss systems is that they encompass a low-intensity farming unit of HNV, which is only an appendix to a more intensive form of land use (the lowland farm unit).

2.5. Baixo Alentejo, Portugal

Two types of landscape characterize the Baixo Alentejo study area: the open field on the flatland (Campo Branco) and, in the surrounding area with a slightly rougher morphology and/or shallower soils, the Montado system. In the open fields, farmers develop a more or less long rotation according with the soil type, based on rain-fed cereal and extensive grazing in fallow areas. This type of land habitat suited to steppe bird species and determined the classification of the area as a Special Conservation Area (ICN, 2006). The Montado system is an agrosilvopastoral system comprising an open formation of cork (Quercus suber L.) and/or holm (Q. rotundifolia Lamk.) oaks, combined with grazing activities (Coelho, 1997; Moreira and Coelho, 1997; Pinto-Correa, 2000; Pinto-Correa and Mascarenhas, 1999).

In the Portuguese case, a proper CLS does not exist, but rather a private-dominant property system with private grazing rights is private-dominant. Main actors are large landowners, either managers of their farm-holding or renters; medium-size farmers who frequently have to rent additional land in support of a mixed crop and livestock operation; and small farmers and landless pastoralists, the latter keeping their animals under renting agree-ments with landowners. Nevertheless, livestock production in this study area is predominantly assured by large landowners even if, in many cases, the wage granted to pastoralists includes the right to keep own animals jointly with the landowner flock.

A more complicated picture appears when land uses and livestock species are considered. In Baixo Alentejo three main land uses are dominant: the open areas of cereal cultivation, where operate a mix of cereal cropping with cattle and sheep grazing; the Montado, open forest of Quercus spp. (holm and cork oaks), where mixed farming of Alentejano pig, meat cattle and sheep may operate with cork extraction; the shrubby encroachment

(11)

CORRECTED

PROO

F

areas under large fallow periods, where hunting appears as a strong compet-itor to grazing activities as well as EU-supported afforestation programs. Tourism operators may overlay in different land uses.

2.6. Castile-La Mancha, Spain

The southern Castilian plain forms most of the basin of the Tajo and

Guadiana rivers. The whole region occupies an area of some 7.8 106_ha

and is divided into five administrative provinces (Albacete, Ciudad Real, Cuenca, Guadalajara, and Toledo) and 916 municipalities. The central part of the region is properly called the La Mancha plain, where arable land is dominant (some 80% of total agricultural land, TAL).

In this Spanish study area, private landownership is dominant with some common grazing land in the mountains surrounding the plain. Mixed arable and sheep operations, where existing, are carried out of the same land units (grazing allotments or polı´gonos de pastos), under private ownership of arable land and public grazing rights, awarded to landless pastoralists (customary use-rights). Sheep farmers (both milk- and meat-oriented) take advantage of agricultural residues in arable land (mainly cereal stubble and fallow land).

3. Material and Methods

Field data were gathered in agreement with common headings and indicators, based on previous coordinate effort (Caballero and Ferna´ndez-Santos, 2004). Most teams except Northern Sapmi, which used public statistics for Norway, used questionnaires as field data collection tool and livestock farmers as individual recipients. In the case of Northern Sapmi, the reindeer district was used as sample unit and official records as source of information gathered at the level of the MU, in this case the husbandry unit within district. In case of the three alpine systems, the whole MU encom-passed two farm units as livestock farmers moved animals from alpine private and commons grazing land over the summer season to farm holdings in the lowland over the rest of the year. In the two Mediterranean systems, the animals stayed over the year in the same MU, being private farm holdings in the Montado of Baixo Alentejo (Portugal) and grazing allot-ments ( polı´gonos de pastos) composed of several farm holdings under public allocation of grazing rights in Castile-La Mancha (Spain).

Questionnaires were drafted based on main criteria as agreed on the matrix-heading: land uses, farm size and land ownership schemes, forage deficit (FD), grazing facilities, stocking, grazing management, economic performance, labor, and institutional factors. Productivity estimates were calculated on either by working unit (WU) or land unit.

(12)

CORRECTED

PROO

F

3.1. Main criteria and indicators

Determining habitat and landscape features that lead to patterns of biodi-versity is an important step for the assessment of the impact of extensifica-tion in agriculture. Local habitat factors for organisms are those influenced by management practices (Jeanneret et al., 2003a) and seminatural biotopes (Zebisch et al., 2004). The first authors stressed the influence of surrounding land use. There are no general models relating overall species diversity to landscape diversity, being the relationship depending on the organism examined. Land uses have been identified in our study areas as a require-ment to assess biodiversity responses. These responses to landscape and habitat changes have to be identified by means of a multiindicator concept in different landscapes situations ( Jeanneret et al., 2003b).

In the six study areas very different kinds of pastoral resources are used, but grazing and nongrazing land uses were differentiated. Intensity of use of grazing resources was also stressed either by accounting the level of use of potential resources or the spatiotemporal distribution of grazing use.

In Northern Sapmi, winter and summer reindeer grounds are differen-tiated, partly within each country and partly across country borders, with Norway having excellent summer pastures in the suboceanic mountain ridge, and Sweden and Finland mostly winter pastures in the dry continental woods. Thus, the annual grazing cycle follows the directions of big river valleys, although borders’ barriers have curtailed the traditional migrations of Sa´mi herders between countries (Riseth et al., 2003).

In the three mountainous areas (Tatra, Entlebuch, and Bavaria), land uses are differentiated by farm units (highland summer pastures and lowland private farms). Grasslands are dominant in both farming units but the degree of use and intensity are different with higher intensity in the lowland farm, especially in Entlebuch and Bavaria, and some risk of abandonment of upland pastures (Grunig et al., 2004). Nongrazing land uses corresponded mainly to alpine forest and some protected areas.

In the two Iberian countries, pastoral resources encompassed a mixture of arable land resources (stubble and fallow land) and natural pastures linked to open oaks’ forest, the first being dominant in Castile-La Mancha and the latter in the Baixo Alentejo. Nongrazing land in these study areas are mainly more intensive cropped areas of vineyards, olives, and plots under irrigation, as well as some protected areas such as subsidized afforestation parcels or young tree plantations.

Farm size and land ownership schemes were defined either in private grazing land or in common grazing land. The latter dominates in the Northern Sapmi study area and in the highland pastures of Bavaria. Private landholdings dominate in Entlebuch, Tatra, in the two Iberian study areas, and in the lowland farms of the three alpine areas. The size of the MU varied largely among study areas, as well as the size of farming units

(13)

CORRECTED

PROO

F

within study areas. Livestock farmers have legal grazing rights in the com-mons of Northern Sapmi and highland pastures of the alpine areas, private grazing rights in all study areas, and consuetudinary (customary) use-rights to arable pastoral resources in Castile-La Mancha. Grazing fees were appli-cable for allocation of grazing rights, either in commons or in private grazing land.

Modeling the FD was defined by a mass balance of available grazing days provided by complementary forage resources (CFR) as compared with structural nongrazing days (Caballero, 1993, 2003). These are days along the year where grazing is hampered by lack of vegetative growth, presence of snow, or humid soils. The only way to avoid an FD under grazing conditions is long- or short-distance migration patterns (trashumancia or trasterminancia). The first migration pattern represents long-distance and horizontal move-ments and the latter, shorter, and vertical movemove-ments. The FD represents the forage coverage of CFR on structural nongrazing season (SNGS).

Most areas of Northern Sapmi have operative long-distance migration patterns though present countries’ borders and grazing restrictions based on international border conventions, to a considerable extent, have shortened or stopped traditional patterns for reindeer herding. Particularly in Finland, with relatively stationary grazing patterns, supplementary feeding covers the FD. Unfortunately, we have no sufficient data to evaluate the implications of this fact.

In the Tatra Mountains and in the two Alps’ study areas, summer grazing days in the highland pastures assures 100–130 grazing days and the potential FD may occur in the lowland farms. In the two latter study areas, the potential FD can be more acute as only a proportion of the livestock units (LU) goes to the highland pastures. Productivity of CFR and proportion of TAL devoted to forage conserves are key issues. In the Iberian study areas, with stationary grazing patterns and climatic constraints, an FD may appear if forage conserves are not provided for coverage of the SNGS.

Across study areas and farm units within study areas, grazing facilities may differ greatly. Fences, barns, water points, milking or slaughtering facilities, haymaking or manure handling machinery, remoteness, accessi-bility paths, or herders’ shelters are important indicators of less hardworking conditions, mobility, and homogeneous grazing use. Grazing facilities are of the outmost importance in highland pastures, remote areas, or grazing units where the pastoralists have limited resources or rights to improve grazing facilities such as in the Tatra Mountains or in Castile-La Mancha.

Stocking was defined as number of LU per hectare of available pastureland over grazing seasons or grazing units. In three of the six study areas, one grazing species is dominant: reindeer in Northern Sapmi, and sheep in the Tatra Mountains and the Castilian plain. In the other three study areas, different livestock species or type of animals are dominants in specific units. In the two alpine areas of Bavaria and Entlebuch, heifers are dominant in the alpine units

(14)

CORRECTED

PROO

F

and dairy cattle in the lowland farms. In Baixo Alentejo, meat cattle is the dominant specie followed by meat sheep, except in the holm oak Montado where the Alentejano pig has its mostly demanded grazing territory. In these three areas with mixed grazing, units of different species and types of animals should be converted to LU by standard tables of equivalence to obtain the size of the herd by grazing unit. Similarly, on each land unit, nongrazing land should be detracted from TAL to obtain available pastureland.

In Northern Sapmi, different stocking and grazing distribution can be related to patterns of migration of reindeer herding. In the Tatra Mountains as well as in the Alps, differential stockings are related to lowland farm stocking, over most of the year, and summer stocking in the highland pastures. In Baixo Alentejo, cereal- or Montado-dominant areas may support differential stocking densities. In the southern Castilian plain, stocking can be related to land uses, either arable or nonarable land-linked resources.

By comparing stocking across study areas and grazing units some insight on grazing distribution can be obtained. The question, however, of whether study areas or specific grazing units are over- or underused remained unchecked. This assessment would require a comparison between potential stocking (base stocking or carrying capacity) and real stocking. Estimation of potential stocking would be based on availability, seasonal productivity, and quality of corresponding pastoral resources by study area or grazing unit. However, the FD mass-balance model, applied to most study areas except Northern Sami, may provide some insight on the adjustment of forage supply to animals’ requirements. In Northern Sapmi, pasture surveys have been used regularly in the latest decades. Several studies indicate considerable overgrazing of lichen resources (particularly winter but to some extent also fall pastures) in Finland and in the Norwegian LACOPE area of western Finnmark as well as the adjacent Karasjok area of eastern Finnmark (Colpaert et al., 2003; Johansen and Karlsen, 1998; Moen and Danell, 2003).

Under the heading of grazing management, mobility of livestock across grazing units was checked within study areas, as well as grazing days within the grazing units. Main schemes of seasonal reproduction (mating/calving-lambing/milking seasons) allowed relating the physiological status of main grazing species with seasonal grazing units. Grazing management also assessed the animal lots under grazing or indoor feeding by grazing season as well as main herding practices. The latter included whether herds/flocks were permanently conducted or only temporarily checked.

Main grazing species, animals’ lots, animals’ breed, production objec-tives, and main indicators of animals’ performance were also recorded in this heading. Different production objectives were recorded across the six study areas and even within one specific study area.

Animals’ performance indicators were required for estimating the value of production farming. For milking lots, marketed milk per dairy cow or per

(15)

CORRECTED

PROO

F

breeding ewe and for meat production, the mean live weight (LW) of animals at selling or slaughter weight in the case of Northern Sapmi were recorded. Some other productivity indicators such as milk production per WU were also recorded. Distribution of milk production over the year or main processing-marketing channels of important livestock products were also recorded in most study areas.

Under this heading, some indicators of future trends in the grazing management of LSGS were also recorded. Current or predicted changes in LSGS management may have some ecological and economic effects. Changes in land uses that can promote better grazing practices and produc-tivity, trends in grazing days, animals’ lots under grazing, or trends in spatial distribution of grazing over the MU were some management indicators recorded. It was also important to assess the trends in the extensive grazing operations in the face of the European debate between extensification and intensification of LSGS (Caravelli, 2000; Marriot et al., 2004; Pinto-Correa and Mascarenhas, 1999; von Boberfeld et al., 2002).

Economic indicators were recorded with the aim of allowing a certain harmonization of reporting and comparison between study areas. Classical cost-benefit analysis was the main tool devised for analysis. The heading was divided in two main tiers: income structure and cost structure. The latter recorded external supply of feeding inputs, animal health expenditure and veterinary assistance, grazing fees, amortization and interests, labor (family or waged), and other costs such as transportation, animal acquisitions, or stock depreciation. Within the income tier, value of production farming, subsidies, and other income sources were recorded. Net profit or losses were calculated by detracting total cost from farming income, either with or without subsidies. Notwithstanding this common economic framework, we opted to maintain the traditional farm accounting of individual study areas instead of looking for a rather artificial harmonization. The main reason for this approach was that we were looking for a general picture of economic sustainability emerging from data of individual study areas rather than cross comparisons of individual study areas. These comparisons are still possible, taking into account particularities of farm accounting.

Taking into account the high degree of heterogeneity of the farm structure between and within study areas, the definition and selection of a meaningful economic indicator is not straightforward. In our study areas we have a gradient of increasing ratio of capital demand to running costs with corresponding increase of imputed costs. The remuneration of the produc-tion factors family labor, own capital, own land, and own assets induce imputed costs. One factor determining the relevance of imputed costs is the productive orientation. Dairy operations, such as in the lowland farms of Entlebuch and Bavaria, have higher imputed costs than the ones focusing on meat production. In these areas, the imputed costs can be in the some order of magnitude as the running expenses.

(16)

CORRECTED

PROO

F

Most of the study areas depend greatly on family labor. Whether a farmer has to take into account this and other imputed costs depends on his/her attitude toward farming and his/her dependence on on-farm income. Only in Baixo Alentejo and Castile-La Mancha, family labor was valued due to the apparent trend of relying on waged labor. To correctly assess the economic sustainability of a given farm, one has to know the personal valuation of the farm-specific production factors. This would allow an adequate assessment of its imputed costs.

Having these problems with the correct assessment of the imputed costs, we selected the cash flow as the main indicator for the cross-country comparison of the economic sustainability. The cash flow has the advantage that it can be traced back to ‘‘real’’ monetary transactions, reducing the potential assessment bias. However, a cash flow of a given magnitude does not imply the same degree of economic sustainability across the study areas. In one system, the imputed costs might be negligible due to the low capital demand of the system and the use of waged labor implying that the cash flow is nearly equal to the profit. In another, the imputed costs might even exceed the running expenses. For the study areas where these considerations play a role, these aspects will be addressed and discussed in the respective para-graphs. In addition, peculiarities of some running or imputed costs in specific study areas are described in the text. With respect to public handouts, the data depict the situation in EU countries before the 2003 CAP reform.

Income and cost tiers were calculated for MU. In some study areas such as in Tatra, Entlebuch, and Bavaria, the whole MU is composed of two farm units: the highland pastures and the lowland farm, with different income, cost, and subsidy tiers attached to the respective unit. In Entlebuch, for example, grazing fees are a source of income instead of cost as owners of the external livestock, added to the alpine unit, pay a grazing fee to the owner/ manager of the alpine unit. In those cases, separated records were available for each farm unit and results can be combined to get a picture of the whole MU. For harmonization of reporting and comparison between study areas, results were recorded for LU of the respective type of animals. In the case of Northern Sapmi, official records were used to assess the economic perfor-mance in Norway, while Sweden and Finland data are mainly based on herder interviews as statistics are incomplete.

Subsidies are an important tier in most extensive European livestock systems. A befuddled complex series of subsidy schemes are operative across study areas and even within one study area. Subsidies are awarded by EU, national and regional governments, or shared by both, and are allocated as direct payments, rural development schemes (agri-environment schemes, less-favored areas, and so on), specific grazing practices, and specific grazing units (a pool of public handouts). Notwithstanding this confusion, total subsidies were recorded by manager/owner of the MU and expressed as percentage of total value of production (OECD, 2001).

(17)

CORRECTED

PROO

F

Under the labor heading, number of WU was recorded per MU on each study area. Labor was differentiated as familiar or waged and as full- or part-time employment. Labor productivity was rated as value of produc-tion farming per WU or number of LU/WU. The working hour (wh) annual standard was 1800 Awh. Working intensity was classified as High (H), Medium (M), and Low (L), regarding care intensity and migration/stationary models of herding.

Legal and institutional frameworks were important especially in those study areas where grazing rights are shared or regulated by some regulatory institution. Entitlement of property rights or renting contracts were also important for grazing regulation or subsidies’ allocation in some areas such as the Tatra Mountain. Government regulations of grazing rights’ and subsidies’ allocation were especially important in study areas dominated by common lands (Northern Sapmi) or landless pastoralists (Castile-La Mancha). Farmers’ opinions were recorded on the sustainability of the legal and institutional framework regulating the grazing operation, recent management trends, as well as the main destabilizing or limiting factors of the grazing systems.

3.2. Management units

The reindeer husbandry matrix was based on the average economic data from the LACOPE target areas in Northern Sapmi. For Finland, only one district or MU was included in LACOPE. Economic data for this district were available. For Sweden, however, only mean regional data of national target areas were available. These regions included several districts: Northern Norrbotten Mountain Sa´mi (nine districts). In Norway, most data were available on district level, here used for western Finnmark (26 pasture districts). LACOPE districts are within these regions with one exception in Sweden. The numbers provided in the data matrix are average for these regions, as individual districts (siida in Norway/Sa´mi village in Sweden) data were not available. In these latter two countries, the MU was the husbandry unit within reindeer districts, where economic data were gathered and most subsidies allocated. National models (three matrices) are based on regions corresponding to each country (Norway, Sweden, and Finland). The national differences were larger in the income tiers than in the cost tiers as a consequence of differences in herder production strategies (Riseth, 2000, 2003), differential prices of reindeer meat, as well as national subsidy systems. In this study area, data were available for the three national models of reindeer herding (Norway, not a member of the EU; Sweden and Finland members of the EU).

Two farming units constitute the grazing system in the Tatra Mountains. The whole MU is thus composed of the lowland farm and the clearing alps. In order to gain a more profound insight in the system, a specific question-naire was designed for each farming unit that took into account their

(18)

CORRECTED

PROO

F

respective peculiarities. As a result of the breeding scheme (one lambing per year), the milking period corresponds to the summer season where tradi-tional cheese making (bundz and oscypek cheeses) takes place. During the lowland farm period, marketed lambs and subsidies are the main sources of income.

In Entlebuch, the whole MU is composed of two farming units, one in the valley as a private farm and one in the alpine pastures (67% as a private unit, 26% as tenant, and 7% in some private or public cooperative agreements).

In Bavaria, the MU also encompassed two farming units. The lowland private farm, where animals stayed over the year. The land is in the ownership of the farmers or rented and is individually exploited. The second type of unit is the cooperative upland Allmende, where farmers sent their heifers for the summer grazing season. These units can be in the ownership of different bodies (private person, local authority, cooperative, and so on) and are jointly used and managed from the lowland private farmers. Although specific incentives are awarded for the use of upland pastures, only some 28% of lowland farmers send the totally of their heifers to the Allmende where they are raised under extensive grazing (Niemeyer and Rosenthal, 2003).

The study area of Baixo Alentejo is the only one where the MU coincides with individual farm holdings.

Individual holdings in Castile-La Mancha, mainly devoted to cereal cultivation, are grouped in large grazing allotments (polı´gonos de pastos) that encompass patches of diversity of resources such as cereal, annual legumes and sunflower stubble, shrubby-steppe vegetation (eriales), natural pastures, and fallow lands. The agricultural land of each municipality is divided, according to its size, into several polı´gonos and each small landowner, having a parcel within the polı´gonos, receive a per hectare grazing-fee paid by landless pastoralists who rented. More than 90% of sheep farmers rely on the polı´gonos de pastos and some of them add small parcels of owned or rented land outside the official system. As arable farming is the primary land use objective and crops are interspersed, the polı´gonos are unfenced, and sheep flocks should be permanently conducted with high working intensity. Individual polı´gonos corresponds to MU in this study area (Caballero, 2001).

3.3. Sampling process

The matrix of data in Northern Sapmi was based on different public and private sources, and encompassed data from national reindeer models in Norway, Sweden, and Finland. The Norwegian reindeer husbandry admin-istration publishes an annual economic report based on numbers and accounts from the reindeer herders and their supervising organizations (konomisk Utvalg, 2004). In Sweden and Finland, there are no annual

(19)

CORRECTED

PROO

F

economic publications. The Swedish data are based on a joint publication issued by the Swedish bureau of statistic and Swedish reindeer-herding organization (SSR, 1999). The Finnish data were received from nonpub-lished sources given by the Association of reindeer herding cooperatives for the year 2003.

Of a total population of 2751 farmers in the Tatra area, 40 lowland sheep farms were sampled. The results for the summer season corresponded to 17 flocks under the care of a main shepherd (baca-unit) and additional labor support by younger shepherds. Data gathered correspond to the year 2003. A main heading-based questionnaire was sent to the 230 owners and/or managers of the alpine units in the Entlebuch Reserve and 107 completed questionnaires were gathered. Effective rate of response to the different tiers of the questionnaire varied from 75% to 100% of the filled questionnaires. Other sources of reported information were used and recorded in the corresponding heading. Most managers and owners of the alpine unit (some 90%) have a farm unit in the valley. Information on these farm units was mostly recorded from BfS (2004) and AfS (2004). Data for both farm units corresponded to the year 2003.

In the prealpine and alpine agrarian regions of Bavaria still exist around 1200 alps, in 155 thereof more than one farmer is involved in their exploi-tation (Allmende). For the purpose of this study, 56 farms participating in CLS and 34 Allmende were surveyed. Of the 56 farms, 38, 13, and 5 are located in the agricultural regions of the Alps, prealps, and prealpine moraine belt, respectively. Of those farms, 43 had entitlements to use the Allmende: 33 of those are located in the Alps and 10 in the prealps. The left 13 farms did not posses any entitlement but board their animals on the CLS. Average farm size increased from 29 LU in the Alps to 62 in the prealps and 82 in the prealpine moraine belt.

In the same way, a number of 34 upland units were investigated. Most of them are located in the alpine region (25), 5 are situated in the prealpine area, 2 in the prealpine moraine belt, and 2 in the southern Bavarian foothills. Mean size of the upland Allmende unit is some 470 ha and ranges from 10 to 7400 ha. The average unit is more or less divided between forest and grazing land. Data for both units corresponded to the year 2003.

In the study area of Baixo Alentejo, 15 mixed-operating farm holdings were sampled. Data corresponded to the agricultural year 2003–2004.

Official records of sheep farmers in Castile-La Mancha, entitled of EU subsidies, amounted to some 8000. In the study area, 231 sheep farmers of the whole region were sampled with the criteria that the 5 provinces were to be represented by at least 5 farmers on each one of the 21 counties in the region. The survey tool was a questionnaire drafted according to main headings and totally 72 variables of quantitative and qualitative character. Surveyed sheep farmers were previously contacted for the local veterinary staff of the Animal Health Associations or Agrupaciones de Defensa Sanitaria

(20)

CORRECTED

PROO

F

(ADS) in its Spanish acronym. The staff concerted working meetings with their corresponding affiliates to explain the objectives of the survey and content of the questionnaire. In this way, the rate of response was almost 100%. Data gathered in this study area corresponded to the year 2002.

4. Results

Typology of grazing systems with policy relevance can be addressed as a combination of analysis-related categories and systemic assessment of com-mon features or trends. Even if a great deal of variation can be found for land use or grazing management indicators, both within or between systems, some common features may arise such as poor economic performance, scarce labor supply, abandonment, consolidation, legal or institutional drawbacks, and poorly devised subsidy schemes. From these main identified issues, policy actions can be derived although proper devising and implementation should be consistent with particularities of the individual systems. The interesting point in comparative typology of our six study areas is to untangle, if existing, these common features in the wide range of variation of most indicators. This report deals mainly with analysis of descriptive categories. In Section 5, however, we will try to uncover common features and trends to the six study areas. Between-system variability will be recorded in tables by indicating study areas’ averages of main indicators, and within-system varia-bility of some significant indicators will be recorded in the text. For those study areas with two farming units per MU (Tatra, Entlebuch, and Bavaria), indicators will be differentiated or weighed either in tables or in the text.

4.1. Land uses

Land uses were related by their potential contribution to the forage supply. In Northern Sapmi and the alpine study areas, nonarable land makes the most significant contribution to pastoral resources, mostly as natural grass-land. In the two Iberian study areas, however, pastoral resources derived from arable land made a significant contribution to the feed supply (Table 2).

In the study area of Northern Sapmi, boreal forest/open tundra and natural alpine grassland dominate while cultivated agricultural lands are limited to valley and fjord areas, in North Norway about 1% of the total land area (Statistics Norway, 2004). Most of grasslands and large proportion of forest/tundra can be used as summer and winter grounds, respectively, for reindeer herding. Imagining a scale from continuous outfields via plots of outfields and managed pastures to plots of infields and indoor feeding, the Northern Sapmi system is still to a very high extent based on feeding from continuous unmanaged pastures.

(21)

CORRECTED

PROO

F

In the Tatra Mountains, land uses recorded in Table 2 corresponded to the lowland farm unit. In the alpine unit (sheep in bacas’ care for the summer grazing season), arable land was underrepresented but pastureland (clearings in the forest) accounted for some 30% of TAL. The rest were alpine forest. In the study area, clearings and forest were in the proportion of 1:1 in hectare.

In Entlebuch, the data recorded corresponded to land in the lowland farm. In the alpine unit, unproductive land and forest take up some 40% of the land, the rest being pastureland of natural grasslands (51%), nature protected areas (5%), and grazing forest (4%).

Similarly, land uses for Bavaria corresponded to the lowland farm unit. In this case, 60% of the farm hectare was composed of intensively managed land (intensive pastures plus arable land) and 20% corresponded to the extensive managed land (litter meadows and alpine pastures). Some 20% of the land managed in the lowland farms were composed of forest.

Referring to the 1999 agricultural census the study area of Baixo Alentejo counted with more than 220,000 ha of TAL of which, 40% corresponded to arable land (temporary crops and fallow), 40% to area under permanent pastures and oak forests, and 20% of shrubland. Broadly, we can consider two different systems. The first, corresponded to cereal growing areas where residues (cereal stubble) are used by cattle and/or sheep and, the second, to the open-forest dominant areas (Montado) where cultivation was only occasional and where a mix of suckle cows, sheep, and Alentejano pigs were operating. Nongrazing land areas included shrubby invaders where only hunting may operate and some nongrazing cropland of vineyards, olives, and parcels under irrigation.

In the southern Castilian plain, arable land was dominant, especially in the central part of the region (La Mancha) where arable land takes up some 80% of TAL. Nonarable land was more significant in the foothill and mountain areas surrounding the plain. In this study area, pastureland included cereal, legumes, and sunflower stubble in the arable land, and natural pastures, eriales (shrub-steppe vegetation), and grazing Mediterra-nean forest in nonarable land. Nongrazing land uses included mainly

Table 2 Land uses and pastoral significance in the six study areas (% TAL)

Indicatora

Northern

Sapmi Tatra Entlebuch Bavaria

Baixo Alentejo Castile-La Mancha Arable land 1 3 0.5 1 40 65 Non-arable Land 99 97 99.5 99 60 35 Pastureland 80 93 99.5 80 75.3 85 a

(22)

CORRECTED

PROO

F

vineyards, olives, and irrigated parcels in the arable land part and dense Mediterranean forest in the nonarable part, encroached by shrubby invaders because of lack of grazing use.

In most study areas, the coefficient of variation (CV) of many land use variables exceeded 0.8 or even 1, indicating very skewed distributions. For instance, in the sample of lowland farms of the Tatra Mountains,

nonarable land per farm was 15 17 ha and area of natural grassland per

farm was 13 16 ha. In the sample of Baixo Alentejo, the mean total

pastured area was 379 322 ha, corresponding to 89% of the total area,

while the proportion of arable land/TAL was 67 72%. In the sample of

Castile-La Mancha, proportion of arable land over TAL was 65 28%

(Table 2), and the proportion of natural pastures plus eriales over TAL was

17 17%. These two and all other land use variables showed asymmetrical

distribution (absence of normality) as rated by the W-test of normality (Shapiro and Wilk, 1965).

4.2. Size of farm-holding, land prices, and grazing fees

In Northern Sapmi, the farm-holding size was of less relevance as reindeer grazing is organized by pasture districts and husbandry units. In the western

Finnmark area (Norway), with 24,290 km2 _{and 241 husbandry units, the}

mean size was some 10,000 ha per husbandry unit (Table 3)

In the Tatra Mountains, the average renting price of pastureland (9€/ha)

and the grazing fee (4 €/ha) corresponded to the lowland farm unit

(Table 3). In the alpine unit (sheep in bacas’ care), mean grazing fees were

4.8€/ha in public lands (Tatra National Park) and 36 €/ha in private lands.

The mean size of the alpine unit was 46 ha/baca flock. The price of land is currently under adaptation to the free market rules but the number of land transactions in the study area of Tatry and Podhale are very limited. Attachment to the land is part of the cultural character and, most frequently, land is transferred within the family. Even land lease is not based on written contracts and even long periods of occupancy do not mean any right for the leaseholder. Land transfer prices are much lower when ‘‘within the family’’

(some 2300€/ha) than for ‘‘outsiders’’ (some 4000 €/ha). In other Carpathian

regions, such as Beskid Niski, prices are much cheaper (some 1000 €/ha),

transactions are more frequent, and do not carry so deep emotions.

In Entlebuch, the mean size of the sampled lowland farm was 14 ha (Table 3) and the mean size of the alpine unit was 57 ha. Managers/owners of the lowland farms do not own enough animals to stock one unit of alpine pastures. They should rely on external animals to stock properly the alpine units. Owners of these external animals pay a grazing fee to alpine owners/managers (Wirz Handbuch, 2004). This side income represented the main income tier (59%), together with subsidies (41%), for the alpine unit operation. For this reason, luring external farmers to bring their animals

(23)

CORRECTED

PROO

F

Table 3 Farm-holding structure, prices, and grazing fees in the study areas Indica to r N or the rn Sa pm i a T at ra Ent leb uch Ba v a ri a Ba ix o Al en te jo a Ca st il e -La Ma n cha Size of the farm holding (ha) 10,000 15 14 37 425 500 Size of herd/flock (LU) 48 12.7 15.6 40.7 141 82 Price of the lowland farm (€ /ha) NA 4000 32,000 25,000 3250 4900 Rent of the lowland farm (€ /ha) NA 9 500 143 50 50 Grazing fees (€ /ha) NA 4 286 22.5 17 3.2 MU using LSGS (%) b 100 90 20 28 NA 90 a In Northern Sapm i, mean size of the husbandry unit in weste rn Finnmark (Norw ay). NA (not applicabl e). In Baixo Alentejo mea n size based on the agricult u re census was 65 ha. However, av erage LSGS involved larger farme rs as it is repre sented in our sa mple (me an size 425 ha). b Mana gement uni ts (MUs ) using coope rative past ures or extensive grazing uni ts (i.e ., alpi ne units in Tatra, Entl ebuch or Bavari a, rented pas tures in Baixo Alentejo or polı ´gonos de pastos in Castile-La Manc ha). Approxi mately, the land price on the alpine unit of Entlebuc h ¼ 10,000 €/ha.

(24)

CORRECTED

PROO

F

to the Alps for summer grazing was of paramount importance for the sustainability of the alpine system. In this case, the grazing fee stated in Table 3 was a mean estimation of external grazing fees and represented a

source of income. Estimation1 _{was based on grazing fee paid by the most}

represented lot (1- to 2-year-old heifers).

In Bavaria, the total hectare of the farms ranged from 3 to 124 ha with a mean size of 37 ha (Table 3). The cooperative upland Allmende ranged from 10 to 7400 ha with an average of 468 ha. The grazing fees in the Allmende were much lower as in the other alpine regions. Mean seasonal stocking (some 120 grazing days) on these units reached 0.9 LU/ha and mean grazing

fee was 25€/LU. Mean grazing fee in the alpine unit (Allmende) was thus

some 22.5€/ha (Table 3). However, it should be stressed that in a lot of

cases (around 50% of the Allmende), no grazing fees were claimed. The other indicators represented in Table 3, for this study area, are mean sizes and prices in the lowland farms.

In the Baixo Alentejo sample, the mean size of farm holdings was

425 256 ha, but only 72 178 ha corresponded to permanent natural

pastures. According with expert knowledge information in all the reference area (Alentejo), it is frequent the acquisitions of grazing rights on a year basis

(or part of the year) with the grazing fees ranging from 10 to 20€/ha per

year according with the quality of the land. In the few cases of our sample where the acquisition of grazing rights was reported, the grazing fees

attained 17.5 €/ha. Renting land (some 9% of TAL) on a yearly basis or

for larger period showed a wide variation, 7€/ha as a minimum to 63 €/ha

as maximum. In Baixo Alentejo study area the price of agriculture land

ranged from 1500 to 5000 €/ha according with the quality of land.

Considering the whole area of the Baixo Alentejo, this range would be enlarged if good clay soils of the Beja area (land price ranging from 4000 to

7500€/ha), or if land located inside the perimeter from the new

Alqueva-dam irrigation system (10,000 to 15,000 €/ha) were taken into

consideration.

In Castile-La Mancha, mean size of farm holdings (500 ha) corresponded to the MU ( polı´gonos de pastos) where individual sheep flocks are main-tained. These MUs are aggregation of individual farm holdings with mean regional size of some 30 ha. In some counties and municipalities, the polı´gono may encompass the landholdings of up to 80 landowner cultivators. Grazing fee corresponded to the sheep allotments under public allocation of grazing rights ( polı´gonos parcelarios). Rented pastureland by private land-owners was two to three times higher, although private landland-owners, who do not own a flock, rarely rent their land for sheep grazing.

1

External grazing fee in Entlebuch¼ 0.9 LU/ha 2.5 heifers/LU 1.27 €/heifer per day 100 days ¼ 286€/ha.

(25)

CORRECTED

PROO

F

Within study area variation for indicators of this heading was also large. In the Tatra Mountains, for example, mean size of farm holdings

was 15 17 ha, and average area of bacas’ flock was 49 13 ha in the

clearings of the alps. In Entlebuch, the altitude of 95 huts of sampled alpine units varied from 900 to 1600 m. Some huts were located in the high alpine area (up to 2500 m), with corresponding variation in size and land uses, land prices, and grazing fees, derived from differences in accessibility. In Bavaria, lowland farm size ranged from 3 to 124 ha in the sample and land

price of agricultural land may reach a maximum of 32,000€/ha (mean of

25,000 €/ha). Although mean size of the Allmende unit is some 470 ha,

considering the nongrazing land (forest and wasteland) it may reach more than 7000 ha. In Castile-La Mancha, mean size of the polı´gonos was

499 513 ha or a CV of more than 100% and mean grazing fees 3.21

3.51€/ha. Mean price of land for selling or renting showed less variation

with values of 4900 1325 €/ha and 50 17 €/ha, respectively.

4.3. Institutional economics

The investigated LSGS were closely linked to specific property rights, especially the ones organized in a collective form (in the following men-tioned as CLS). The CLS studied manifest a number of institutional features. First, to some extent, CLS have accommodated a certain welfare institution within their own institutional limits by providing livelihood security to people with very limited alternative possibilities. Second, CLS provided access equity and conflicts resolution for its participants as a functional necessity. Third, there are complex relations between the institutional system and the mode of production including embedded cultural features making the production system viable. Fourth, CLS by mostly being based on some form of rotational and limited use of pastures contributed to resource preservation and ecological sustainability. Concluding on institu-tional properties, CLS had much in common with common-pool resources (CPR). The users have to make collective agreements and have to decide how the resource use can be arranged in such a fashion so that the benefit of each user is proportional to the effort of that user. Moreover, CLS had in cases served as a vehicle for the social distribution of goods among the deprived segments of the population and thus had a potential to contribute as a buffer to take care of the destitute parts of a population.

A comparative analysis of the organization and structural form of the grazing systems presented the role different groups of actors play in these systems and their interrelationships (Gueydon et al., 2004). We could distinguish four main groups of roles: the landowners, resource owners, livestock owners, and pastoralists. The role of the landowner was to provide a part or all of the pastoral resource. The role of the resource owner was to hold the right to exploit a part or all of the resource. The role of the

(26)

CORRECTED

PROO

F

livestock owner was characterized by the fact that he/she owns some or all the animals grazing on the resource. Finally, the role of the pastoralist was to conduct on his own or together with others the herds on the resource.

As shown in Table 4 the landowner can be a single entity for example the state such as in Northern Sapmi and Bavaria. It can be several individual persons acting independently like in the Tatra, in Baixo Alentejo, and also in some cases in Northern Sapmi and Bavaria. It can finally be a community of landowners like in Castile-La Mancha and sometimes in Bavaria or a formal legal entity like in Entlebuch and sometimes in Bavaria.

In Northern Sapmi, Castile-La Mancha, and Entlebuch, and in the more frequent setting of Bavaria, community of people (landowners or livestock owners) jointly owned the resources. In Baixo Alentejo and in the Tatra Mountains the landowners were likewise the owners of the resource. They rent the land or sell the resource for grazing activities under market condi-tions. The land and resource’s ownership was separated in Northern Sapmi, Castile-La Mancha, and in some cases in Bavaria. The land belongs to individual landowners or the state but these entities do not have any statement to issue concerning the utilization or the distribution of the pastoral resource. In most cases the landowners received only limited revenues, if at all, for contributing their land to the system.

In all cases, livestock was individually owned, implying that profits from the exploitation of the resource by selling marketable products, receiving subsidies related to the number and kind of livestock or related to the way the grazing is performed, were not shared.

In most cases, except in Bavaria and Northern Sapmi, the direct utiliza-tion of the resource was under a single appropriator. Generally in cases of a single appropriator, the pastoralist was one of the livestock owners who may board animals of other livestock owners on his own account (Tatra and Entlebuch) or was being paid a fixed wage by livestock owners (alpine areas of Bavaria and sometimes in Entlebuch). In Baixo Alentejo, it was also frequent that the herdsmen combine a fix wage with the right to freely graze their own animals. In the prealpine region of Bavaria the livestock owners were also pastoralists as the work requirements for taking care of the livestock do not demand the employment of a herdsman. In Northern Sapmi, the pastoral-related work like herding and preparation of the herd for slaughtering was done cooperatively by the Sa´mi. In the Tatra Moun-tains, the pastoralist can be landowner and therefore also one of the resource owners and he frequently owns a significant part of the herded livestock himself.

The clarification of the different groups of actors and their role gave indication on the action which are collectively undertaken and conse-quently help to systemize the notion of ‘‘cooperative systems’’ within the different regions. Two different types of activities were carried out together; these were the collective provision of the land and the collective utilization

(27)

CORRECTED

PROO

F

Table 4 Actors inv olv ed in the cooperativ e liv estock systems (CLS) A ct ors N o rthe rn Sa pm i T at ra M ountai ns Ent leb uch Ba v a ri a Bai x o A le n tejo Cas ti lla-La Ma n cha Landowner (resource provider) Individual landowners Individual landowners Community of landowners (formal) Community of landowners (informal or formal) Individual landowners (Community of) small landowners State State Collective private body a Individual landowner Resource owner Joint ownership of Sa ´mi (siida ) bor reindeer pasture ‘‘district’’ Individual landowner Joint ownership of landowners Joint ownership of landowners Landowner Joint ownership of sheep holders and landowners Joint ownership of livestock owners Community of right holders Livestock owner Individual livestock owner Pastoralist (resource user) Jointly the members of siida Individual pastoralist Individual pastoralist Same as ‘‘resource owner’’ (or herdsmen) Individual pastoralists Individual pastoralist a Finnmark State and preliminary proper ty during proces s o f land reform (San dberg, 2006), Norw ay; crown/s tate in Swede n/Fin land but Sa ´mi land claim s. b Nor w egian expression. In Sw ed en it is the S ami v illag e and in F inl and Pa lisku n ta (cooperativ e). B u t in essence it is the same for m of o rgan iz atio n amo n g the th ree cou nt ries.

(28)

CORRECTED

PROO

F

of the resources. Table 5 gives an overview of the presence of these actions in the different grazing systems. One major reason to opt for collective action was to realize economies of scale or to reduce transaction costs (Gueydon et al., 2004).

The degree of governmental involvement varied significantly between the different study areas. In Northern Sapmi, Castile-La Mancha, and Entlebuch, the systems were strongly regulated by external rules and specific public laws. In contrast, the authorities in the Tatra Mountains, Baixo Alentejo, and Bavaria were not involved in the management of the sys-tems apart from general regulations dealing with ‘‘good agricultural prac-tices.’’ These rules applied to all farmers or in the case of Baixo Alentejo to the farmers benefiting from the Zonal Plan of Castro Verde, which is an EU agri-environmental measure aimed at the preservation of steppe birds such as the Great Bustard (Otis tarda). Moreover, the investigated systems ranged from ones with a relatively rigid internal structure and rule system, like in Entlebuch and Bavaria, to others with a high degree of governmental involvement, like in Castile-La Mancha and in Northern Sapmi.

Table 5 Organization forms and collective actions in european grazing systems

Study area Organization forms

Collective provision of the land Collective utilization of the resource

N. Fennoscandia Sa´mi pasture ‘‘district’’ (formal) Siida (cultural origin)

No Yes

Tatra Mountains Private property of the

alpine meadows

No Yes

Entlebuch Private property under

private or cooperative law

No (Yes)

Bavaria Allmende (Yes) Yes